data sharing:
social and normative
kaitlin thaney
program manager, science commons
chantilly, va - ISWC - 25 oct 2009
This presentation is licensed under the CreativeCommons-Attribution-3.0 license.

make sharing easy, legal and scalable
integrated approach
building part of the infrastructure for
knowledge sharing

I am not a lawyer.
(first things first)

social and normative issues
human involvement, added roadblocks
implications of FLOSS
interoperability, design decisions
how to navigate?

the data web

needs to be legally and technically
accessible

“ By open access to the literature, we mean its free
availability on the public internet, permitting users
to read, download, copy, distribute, print, search, or link
to the full texts of the articles, crawl them for
indexing, pass them as data to software, or use them for
any other lawful purpose, without financial, legal or
technical barriers other than those inseparable from
gaining access to the internet itself.”
Image from the Public Library of Science, licensed to the public, under
CC-BY-3.0

as a means to achieve Open Access
but what about data?

knowledge?
journal articles
data
ontologies
annotations
plasmids and cell lines

knowledge?
journal articles
data
ontologies
annotations
plasmids and cell lines
... how to treat? like content? software?

early days of WWW
no licenses (even free)
debate over code
CERN’s decision
view/edit source
network effects

the data web
still in its infancy ...

“the future is here ...
just unevenly distributed”
- william gibson
(i.e., linked data, W3C, neurocommons...)

(social) implications
of FLOSS toggles

free/libre license ethos
notion of licensing to make
more free

©
“creative expression”

is it creative?

is it creative?

is it creative?

category errors

the problem of...
Non-Commercial
for data

Non-Commercial
what’s a commercial use
of the data web?

the problem of...
Share Alike
for data

issue of license proliferation
whatever you do to the least of the
databases, you do to the integrated system
(the most restrictive wins)
risk for unintended consequences

the problem of...
Attribution
for data

social aspect of semantics

agreement
is hard.

espresso
coffee
cafe
kopi
cafezinho
latte koffee
mocha americano

“choice” or interoperability.
(pick one)

converge on common names
“coffee”
“cafe” coffee
“kopi” http://ontology.foo.org/1234567

national law hurdles
sui generis,
“sweat of the brow”
Crown copyright
“level of skill”
how internat’l data sharing efforts
are affected?

protection instinct / culture of control
quality control, integrity concerns
“my data”, interpretation issues
fear, uncertainty, doubt (FUD)

<mosquitos><transmit><malaria>
validation, provenance
relationship mapping, citation?
what rights?
still not fully automated

a norms approach

a non-license means to request
certain behavior
community norms
best practices, terms of use

attribution vs. citation
which one applies? which is best fit?
what’s the difference?
“credit where credit is due”

attribution
“the requirement to acknowledge or
credit the author of a work which is
used or appears in another work”
citation
“reference to a published or
unpublished source” ... prime purpose is
of “intellectual honesty”
(via wikipedia)

attribution:
(legal entity)
“triggered by making of a copy”
does it apply to facts?
how? (papers, ontologies, data)
“in a manner specified by ...”
attribution stacking

how to perform?
how much is enough?
unintentional infringement
(example, ontologies)
does it apply?

citation:
(gentle(wo)man’s club)
legal requirement?
interoperability?
credit where credit is due
entrenched scientific norm

we shouldn’t use the law to make it
hard to do the wrong thing ...

need for a legally accurate and
simple solution
reducing or eliminating the need to make
the distinction of what’s protected
requires modular, standards based
approach to licensing

calls for data providers to waive all rights
necessary for data extraction and re-use
requires provider place no additional
obligations (like share-alike) to limit
downstream use
request behavior (like attribution) through
norms and terms of use

... must promote legal predictability and certainty.
... must be easy to use and understand.
... must impose the lowest possible transaction costs on
users.
full text:
http://sciencecommons.org/projects/publishing/open-access-data-protocol/

set of principles (not license)
open, accessible, interoperable
know it’s safe to play

impose “toggles” through norms,
terms of use
best fit for the discipline
doesn’t limit downstream use

at best, we’re partially right.
at worst, we’re really wrong.

resist the temptation to treat
as property
embrace the potential to treat instead
as a network resource

scalability is key
“get law out of the way”
build + allow for network effects

the right to fix our mistakes.

(remember Prodigy and AOL?)

thank you.
kaitlin@creativecommons.org
sciencecommons.org
creativecommons.org
slideshare.net/kaythaney